JP2003532895A - Method and apparatus for non-destructive evaluation of sputter target cleanliness using detection of acoustic wave phase change - Google Patents

Abstract

Abstract: An improved method and apparatus for non-destructive cleanliness assessment of sputter targets using high frequency waveform phase change and amplitude detection is disclosed. The apparatus acquires phase changes and amplitudes at a plurality of data points. The disclosed method for characterizing a sputter target material (52) uses a phase change and amplitude intensity data to calculate a cleanliness factor and creates a Pareto histogram.

Description

Detailed Description of the Invention

The present invention relates to non-destructive testing methods and apparatus for identifying the types of intrinsic defects in metal sputter target materials. More particularly, it relates to a non-destructive method and apparatus for the identification and counting of solid inclusions using the detection of phase changes in high frequency echo waveforms.

Cathode sputtering is widely used to deposit thin layers or films of material from sputter targets onto the required substrates, such as semiconductor wafers. Basically, a cathode assembly containing a sputter target is placed with an anode in a chamber filled with an inert gas, preferably argon. The required substrate is placed in the chamber near the anode and the receiving surface is oriented perpendicular to the path between the cathode assembly and the anode. A high piezoelectric field is applied between the cathode assembly and the anode.

Electrons emitted from the cathode assembly ionize the inert gas. next,
The positively charged ions of the inert gas are accelerated by the electric field toward the sputtering surface of the sputter target. The material knocked out of the sputter target by ion bombardment traverses the chamber and adheres to the receiving surface of the substrate, producing a thin layer or film.

One factor affecting the quality of the layer or film produced by the sputtering process is the "cleanness" of the material from which the sputter target is manufactured. The term "cleanliness" is widely used in the semiconductor industry, and in particular to characterize high and ultra-high purity materials. In the usual practice, "cleanliness" refers to the internal purity of a material. Impurities can be present, for example, as trace amounts of foreign elements in dispersed or localized form in the sputter target material. Cleanliness is usually the number of atoms per million (parts) that defines the ratio of the number of impurity atoms to the total number of sampled atoms.
It is measured in terms of the number of icles) (“ppm”) or the number of atoms per billion (“ppb”).

Clearly, it is desirable to use relatively clean materials in the manufacture of sputter targets, as the cleanliness of the material from which the sputter target is manufactured affects the quality of the layer or film produced by the target. . This means to those skilled in the art that there is a need for non-destructive inspection techniques for the selection of clean sputter target blanks suitable for the production of high quality sputter targets. Known destructive inspection methods such as glow discharge mass spectrometry and LECO technology are not suitable for this purpose.

Another factor affecting the quality of the layer or film produced by the spattering process is the presence of "defects" in the sputter target material. As used herein, the term "defect" refers to microvolume defects present in the sputter target material, such as inclusions, pores, cavities, and micro-lamination.
means n). However, not all defects are "similar" in terms of degrading effects on sputter performance. Some defects, such as micro-cavities, or shrinking porosity, have a relatively "moderate" degrading effect on the sputter performance, while other defects, such as dielectric inclusions, can seriously interfere with the sputter process. Indicates. Therefore, those skilled in the art need a non-destructive method of identifying and individually counting the various types of defects that may be present in a sputter target material.

FIG. 1 illustrates a prior art non-destructive ultrasonic “defect” detection method for characterizing aluminum and aluminum alloy sputter target materials. The method shown in FIG. 1 is based on Aluminium Pechney's PCT patent application PCT / FR.
96/01959, internal "decohesi" detected per unit volume of an aluminum or aluminum alloy blank.
The size and number of "on)" are similar to those proposed for use in classifying blanks suitable for the production of sputter targets.

The prior art method of FIG. 1 uses a pulse echo method performed on a test sample 10 having a flat top surface 12 and a flat bottom surface 14 parallel thereto. In this method, a focused ultrasonic transducer 16 moves a series of positions on the upper surface 12 of the test sample 10 at least 5 MHz, preferably 10 MHz.
Single short duration, high frequency ultrasonic pulse 1 with a frequency of ~ 50 MHz
Irradiate with 8. The ultrasonic transducer 16 is then switched to the detection mode and detects the series of echoes 20 induced by the ultrasonic pulse 18.

One of the contributing factors to these echoes 20 is the scattering of acoustic energy from the ultrasonic pulses 18 by the defects 22 in the test sample 10. Compare the amplitude of the echo induced in the test sample 10 with the amplitude of the echo induced in a reference sample (not shown) having a composition similar to that of the test sample and a flat bottom blind hole of constant depth and diameter. Thus, the defect 22 in the test sample 10 can be detected and counted.

The number of defects detected by the method of FIG. 1 must be normalized to facilitate comparison of test samples of various sizes and shapes. Usually, the number of defects is normalized by volume. That is, the sputter target material is characterized in terms of "number of defects per cubic centimeter". The volume belonging to the echo 20 from each irradiation of the test sample 10 is approximately 1 pulse within the test sample 10.
It is determined by evaluating the effective area of 8.

A part of the scattered energy is attenuated by the substance forming the test sample 10. Furthermore, the dimensions of each of the defects in question are in the range of about 0.04 to 0.8 mm, comparable to the wavelength of the ultrasonic waves in the metal (e.g. for the frequency range 10 MHz to 50 MHz). The wavelength of sound waves in aluminum is from 0.6 mm to 0.12
mm), thus the pulse 18 has a tendency to refract around the defect 22, which reduces the scattering intensity.

Another factor that impairs the ability of transducer 16 to detect acoustic energy scattered by defects 22 is the physical properties of the defective material, or more precisely, the acoustic impedance at the defect-matrix material boundary. The size of the mismatch. This impedance mismatch directly affects the reflection and transmission characteristics of ultrasonic waves at the phase boundaries. The reflection coefficient of the ultrasonic beam at the matrix-defect boundary can be represented by the simple equation R = (I ₂ −I ₁ ) / (I ₂ + I ₁ ). In this equation, I ₂ is the acoustic impedance of the defective material and I ₁ is the acoustic impedance of the matrix material. A few important conclusions can be drawn from a brief discussion of this equation. First, if the acoustic impedance I ₂ of the defect is less than the acoustic impedance I ₁ of the matrix I ₁ , the R coefficient will be negative. A negative R is interpreted as a 180 ° phase change of the acoustic pulse waveform. For example, the defect has an acoustic impedance of 0.93 g / (cm ² · sec) (× 1
0 ⁶ ) (air) or less (vacuum), gas filled or vacuum (contraction) voids, the phase of the ultrasonic pulse waveform changes by 180 ° at the boundary. Second, the defect has an acoustic impedance of 17.2 g / (cm ² · sec) (
If the gas in the aluminum matrix is × 10 ⁶ ) or the void is a vacuum, the reflection coefficient value is close to the unit or 100%, and the amplitude of the reflected signal is between the defect size and the ultrasonic beam focal spot size. It is a function of only the relationship between. Thirdly, defects solid particles, for example, aluminum matrix acoustic impedance ^{(17.2g / (cm 2 · sec} ) (× 10 6)) of the acoustic impedance is more than twice 39.6g / (cm ² · sec ) (× 10 ⁶ ), the ultrasonic waveform does not cause phase inversion at the defect boundary, and in the case of alumina inclusion, the reflection coefficient does not exceed 39.5% of the amplitude of the incident pulse. (When not considering the wave interference effect). In this case, the amplitude of the reflected signal is a function of two variables: firstly the relationship between the defect size and the beam focal spot size, and secondly the size of the acoustic impedance mismatch at the defect-matrix boundary. Is a function.

Therefore, the final conclusion is that void defects and alumina inclusions of the same size reflect ultrasonic energy in very different ways. In addition to the inversion of the waveform phase, the reflected signal from the void-shaped defect has at least twice the amplitude of the reflected signal from the alumina particle inclusions. Therefore, the detection capability of alumina inclusions is generally smaller than the detection capability of void-like defects, and when the phase information of the reflected signal cannot be extracted simultaneously with the amplitude information, the test result is
Mistaking what is in fact large alumina particles for small void-like defects and vice versa can be misleading.

Transducer 1 for detecting acoustic energy scattered by defects 22
Another factor that compromises the ability of 6 is the noise generated by the scattering of pulses 18 at the boundaries between particles with different textures. In fact
The texture-related noise related to this background tissue is several mm.
In the case of high-purity aluminum having a particle size of about 0, it is very large, and therefore about 0
． It can be said that small defects in the size range of 05 mm or less cannot be detected. When the grain size is large, the signal-to-noise ratio in the acoustic energy scattered by the defects is reduced due to the noise induced by grain boundaries.

Other factors affecting the sensitivity and resolution of the method of FIG. 1 include pulse frequency,
There are durations and waveforms, beam focus extents and focal spot sizes, coupling conditions (ie, the efficiency with which acoustic energy travels from transducer 16 to test sample 10), and data acquisition system parameters.

One major drawback of the method of FIG. 1 is the inability to distinguish between different types of defects, in particular void-like defects and solid particle inclusions such as alumina particles. This method, which relies on echo amplitude measurements, only confirms the physical presence of defects. The physical properties and actual dimensions of the defects are not evaluated correctly and can only be inferred based on the defect type assumptions. Internal “Decohesion”
If the (void-like defect) is the only defect in the target material, this method (Fig. 1
The techniques used in ()) are sufficient to detect defects and to determine dimensions. But in practice, the internal "decohation" referred to in this way (Fig. 1) is
It is part of several defect types that may be present in the target material. For example, metallographic examination also reveals aluminum oxide particles in the aluminum for the sputter target. Therefore, the technique used in this method (FIG. 1) cannot distinguish and distinguish between multiple defect types. This is because the waveform phase change information is overlooked.

Accordingly, there is still a need in the art for a non-destructive method for characterizing sputter target materials having multiple defect types. Also, certain groups of defects such as void defects (cavities, microlayers, “
There is also a need for techniques to individually compare target internal volume cleanliness with respect to decohesion ") and solid inclusions.

Sonix, Inc. (8700 Morrisette Dr., Spri
ngfield, VA22152) is a matrix observation in C-scanning, which is one image observation method implemented by FlexSCAN-C.
A phase gating method is used to detect the phase inversion of the waveform at the defect boundary. This method uses the "Texas Instruments" phase inversion algorithm (licensed to SONIX, Inc.). In this method, a defect is mapped onto a two-dimensional sample image only when a phase change of 180 ° is detected. Therefore, this method is limited to the detection and mapping of void defects when the impedance changes from large to small at the defect boundaries. However, when it is used as a sputter target, it is absolutely necessary to detect and identify alumina particle inclusions. The phase inversion method used by does not work in this case. This is because the waveform does not change its phase at the boundary of this defect.

There is also a need for a method of individually detecting and sizing each alumina particle inclusion.

These requirements and others are non-destructive methods for the characterization of sputter target materials, wherein the test sample is sequentially exposed to acoustic energy at multiple locations on the surface of the test sample of the sputter target material. Then, the echo induced by the acoustic energy is detected, the backscattering noise related to the background tissue is distinguished from the echo, and an uncorrected high-frequency echo waveform signal is obtained. Monitor for phase inversion and compare the uncorrected waveform signal with at least one of each of the phase inversion reference value and the phase non-inversion reference value to obtain void and particle defect data points separately, and Data points are obtained and the defect data points corresponding to each defect type are counted individually and at the same time the total is counted to give a total defect count. Number C _{FT (total)} , C _{FI (with phase reversal)} , phase non-reversal defect count number C _{FN (without phase reversal)} is determined, and defect data points and non-defect data points are counted, and the total number of data points C _DP The total cleanliness factor F _CT = (C _FT / C _DP ) × 10 ⁶ and the individual cleanliness factor F _CI = (C _FI / C _DP ) × 10 ⁶ for each type of defect, and F Calculating _C = (C _FN / C _DP ) × 10 ⁶ , characterized in that it consists of steps.

Unlike the prior art methods described above, the method of the present invention characterizes the sputter target material by separately identifying and counting void-like defects and particle-like defects. Separating the cleanliness factor into components that correspond to different types of defects refines the failure criteria to be more accurate by identifying and sizing different types of defects.

Sonix, Inc. Unlike the method described above, the method of the present invention characterizes both waveform phase inversion defects and phase non-inversion defects. Therefore, the risk of missing the first important non-waveform phase inversion defect in sputter target applications is reduced.

Although the cleanliness factor technique is effective in characterizing sputter target materials, histograms can provide additional insight. More specifically, this sputter target test method defines a plurality of amplitude bands for each type of defect (waveform inversion and non-inversion), measures the modified amplitude signal to determine a modified amplitude signal strength, and Comparing amplitude signal strengths to the plurality of amplitude bands to define a subset of the strengths of the modified amplitude signal, counting the subset of modified amplitude signal strengths to obtain a plurality of modified amplitude signal counts, Each corrected amplitude signal count number of the plurality of amplitude signal count numbers is made to correspond to one of the plurality of amplitude bands, one Pareto histogram is created by combining individual Pareto histograms for both defect types, and the corrected signal It can be said that the method comprises each step of relating a count number to the plurality of amplitude bands. This histogram does not attempt to directly map the location of defects along the surface of the sputter target material and is therefore immune to scaling issues.

Most preferably, the test sample is compressed in the direction of one dimension, for example by rolling or forging, and then propagated generally along that dimension (ie obliquely to the surface or more preferably perpendicular to the surface). Irradiate with acoustic energy. This compaction has the additional effect of flattening and spreading certain defects (aluminum oxide thin film clusters and voids) in the material. This widening of the defects increases the intensity of the acoustic energy scattered by the defects and reduces the likelihood that the acoustic energy will be refracted around the defects.

These methods of determining the properties of sputter target materials can be used in the process of making sputter targets. As mentioned above, the cleanliness of a sputter target, especially with respect to non-phase inversion defects, is a major factor in determining the quality of the layer or film produced by that target. High quality sputter targets are produced by molding only sputter target blanks with cleanliness factors or histograms that meet certain criteria for making sputter targets and rejecting blanks that do not meet such criteria. The likelihood of producing a layer or film of is increased.

Therefore, one object of the present invention is to provide a non-destructive method for the characterization of sputter target materials. Other objects of the invention will be apparent from the following description, the accompanying drawings and the claims.

FIG. 3 illustrates a particularly preferred method for classifying defects and determining the cleanliness of sputter target materials. In this method, a cylindrical sample 50 of sputter target material (preferably made of a metal or metal alloy) is compressed or processed to produce a disc-shaped test sample 52. The sample 52 has a flat top surface 54 and a flat bottom surface 56, the bottom surface 56 being substantially parallel to the top surface 54. A focused ultrasonic transducer 60 is then placed near the upper surface 54. The transducer 60 includes the upper surface 5 of the test sample 52.
4, single, short duration, megahertz frequency range acoustic energy pulse 62
Irradiate. The transducer 60 then detects the echo 64 induced in the test sample 52 by the acoustic energy pulse 62. Transducer 6
0 converts this echo into a high frequency electrical signal (not shown), processes this electrical signal and retrieves information on the waveform phase and maximum amplitude.

More specifically, the sample 50 is first compressed in the direction of dimension 70, as shown in FIG. 2, to produce a disc-shaped test sample 52 as shown in FIG. Preferably, sample 50 is compressed by forging or rolling of sample 50, followed by diamond cutter cutting to form flat surfaces 54 and 56. Dimension 70 reduction is 0 to 1
It can be any value within the range of 00%. The compression of sample 50 flattens and widens some defects 72, thus increasing the defect surface area perpendicular to dimension 70.

As shown in FIG. 4, the test sample 52 is then immersed in deionized water (not shown) in a conventional immersion tank 80. The transducer 60 is attached to a mechanical XY scanner 82, which is electrically connected to a controller 84, such as a PC controller. The controller 84 controls the mechanical X-Y scanning device 82 to operate the test sample 52.
It is programmed in the usual manner to move the transducer 60 in a stepwise motion such as a raster scan across the upper surface 54.

Returning to FIG. 3, the preferred transducer 60 herein is ULTRAN U
SA is commercially available under the name WS50-10P4.5. This is a long focal length piezoelectric transducer with a fixed focal length of 114 mm (in water). At a peak frequency of about 9.15 MHz with a bandwidth of 8 Hz (-6 dB), the transducer has a focal area of about 21 mm in aluminum.
zone (-6 dB) and a pulse of acoustic energy 62 having a focal spot of 0.8-0.9 mm in diameter.

Most preferably, the upper surface 54 of the sample 52 has a width or diameter on the order of about 28 cm. With a data acquisition step of about 0.9 mm in both the "x" and "y" directions, there is no overlap of exposed areas and a detection level of 0.
It is possible to detect flat bottom holes of 25 mm. In this case, about 100,0 on the upper surface 54
The irradiation of the test point of 00 is performed.

Most preferably, the transducer 60 has an acoustic energy pulse 62.
Are propagated in deionized water (not shown) in dip tank 80 and oriented so as to strike test sample 52 approximately perpendicular to top surface 54. Further, the transducer 60 preferably has acoustic energy pulses 62 about 3-24 below the upper surface 54.
It is located away from the upper surface 54 so as to focus the zone 86 of the test sample 52 in mm. The acoustic energy pulse 62 interacts with the sample 52 to induce an echo 64, which propagates in deionized water (not shown), about 60 μsec after the acoustic energy pulse is delivered, the transducer. Return to 60.

Returning to FIG. 4, a particularly preferred echo acquisition system is a low noise gated (
gated) a preamplifier 90, a low noise linear amplifier 92 with a set of calibrated attenuators, a 12-bit (2.44 mV / bit) AD converter 94, and a digital oscilloscope 95 connected to the receiver analog output. Have. When enough time has passed for the echo to reach the transducer 60, the controller 84 switches the transducer 60 from the transmit mode to the gated electronic receive mode. The echo 64 is received by the transducer 60 and converted into an RF electrical signal (not shown). This RF signal is a preamplifier 90 and a low noise linear amplifier 9
Amplified by 2, a corrected amplitude signal is generated, displayed on the screen of the oscilloscope 95, and waveform phase information is extracted. The modified amplitude signal is then digitized by the AD converter 94 before being sent to the controller 84. This AD conversion is performed so as to preserve the amplitude information from the analog modified amplitude signal.

The determination of the nature of each defect (voids or alumina inclusions) is made by monitoring the waveform phase inversion with a digital oscilloscope 95. The size of each defect was determined by a digitized modified amplitude signal obtained from sample 52 intentionally embedded in a flat bottom blind hole or reference sample material of approximately the same composition and constant depth and diameter as test sample 10. Reference sample with alumina particles of size (
It is detected by comparison with a reference value (or calibration value) derived from tests carried out on (not shown).

A particularly preferred PC controller 84 controls the data acquisition process. A particularly preferred software package for use with this data acquisition system is S
structural Diagnostics, Inc. From SDI-531
1 You can purchase the one named Winscan.

The PC controller 84 is also programmed to calculate the total cleanliness factor of the material of the samples 50, 52 and each cleanliness factor characteristic of the classified defect. Rather, it is programmed to identify backscatter noise due to background tissue and further distinguish "void defect data points from alumina particle defect data points". The PC controller 84 continues counting defect data points detected during testing of the test sample 52 and determines defect counts "C _FT ", "C _FI ", "C _FN ".

The PC controller 84 is also programmed to identify “defect free data points”. A "defect free data point" is a digitized modified amplitude signal having an amplitude that, when compared to a reference value, indicates the absence of a defect.

The PC controller also determines the total number of data points “C _DP ”, ie the sum of the defect count number C _FT and the number of defect-free data points. The total number of data points may be determined by continuing to count defective and non-defective data points, but preferably the total number of positions "C" along the upper surface 54 where the test sample 52 is illuminated by the transducer 60. It is determined by counting _I "and subtracting the number" C _N "of digitized RF signals. “C _N ” is the number of signals that the data acquisition circuit cannot determine, due to noise or other causes, to be defective or non-defective data points. (Alternatively, the "noise count number" C _N can be said to be the number of positions along the upper surface 54 where neither defective or non-defective data points are detected.)

The PC controller determines the cleanliness factor F _CT = (C _FT / C _DP ) × 10 ⁶ , F _CI after determining the defect counts C _FT , C _FI , C _FN and the total number of data points C _DP. = (C _FI / C _{D P} ) × 10 ⁶ , F _CN = (C _FN / C _DP ) × 10 ^6, and is programmed to characterize the materials that make up samples 50, 52. Unlike the "defects per cubic centimeter" of the prior art, the magnitude of this cleanliness factor does not depend on the pulse cross-section estimate. This cleanliness factor is normalized by the dimensionless coefficient C _DP × 10 ^-6 rather than by volume, so it is more closely related to the units of ppm and ppb as compared to the unit of "defects per cubic centimeter". There is.

The creation of a program suitable for determining the cleanliness factor according to the invention disclosed herein is within the ordinary skill of those in the art and does not require unreasonably high experimental skill.

Another way to characterize the materials that make up samples 50, 52 is by determining the size distribution of defects within test sample 52. More particularly, the amplitude band or range is defined and the amplitude represented by the digitized modified amplitude signal for some types of defects (phase inversion and phase inversion) is compared to the amplitude band to compare the amplitude of the modified amplitude signal. A Pareto defining a subset, counting these subsets of the corrected amplitude signal to determine a corrected amplitude signal count for each amplitude band and each type of defect, and relating the corrected signal count to the plurality of amplitude bands. By configuring the histogram, the cleanliness characteristics of sample 52 can be characterized. This histogram is an indication of the defect size distribution within the sample 52 because the amplitude represented by the digitized modified amplitude signal for each type of defect is related to the size of the defect detected in the sample 52.

FIGS. 5 and 6 show the characteristics of two Al-0.5 wt% Cu alloy sputter target materials having an orthorhombic structure and a crystal grain size in the range of 0.08 to 0.12 mm. Shows a histogram. The material of FIG. 5 is "cleaner" than the material of FIG. The material of FIG. 5 has a cleanliness factor C _{FT of} 250 and a cleanliness factor C _FN of 100, while the material of FIG. 6 has a cleanliness factor C _FT of 1,200 and a cleanliness factor C _FN of 300. It is important to emphasize that having a small C _FN value is more important than having a small C _FT value for sputtering applications. The defect observation area is within the gate having a duration of 7 μsec and the gate delay is 1 μsec.

The horizontal axis 155 of the Pareto histogram in FIG. 5 shows the amplitude, normalized as a percentage, of the echo amplitude induced in a reference sample with a 0.8 mm flat bottom blind hole. The vertical axis 157 in FIG. 5 indicates the number of corrected signal counts at each amplitude on a logarithmic scale.
The echo amplitude threshold in this defect count was set to 12%. It was confirmed by experiments that the echo amplitude due to the background structure did not exceed 10% in all the tested aluminum alloys. Similarly, the horizontal axis 161 and the vertical axis 163 of the histogram of FIG. 6 are also scaled.

The histograms of FIGS. 4 and 5 are prior art image viewing methods in that the distribution of defect sizes can be represented without the need to display defect sizes (not shown) relative to the surface area of the test sample. Better than

The creation of a program suitable for plotting histograms such as those shown in FIGS. 5 and 6 in accordance with the invention disclosed herein is within the ordinary skill of those in the art and requires unreasonably high experimental skill. do not do.

The cleanliness factor or histograms as shown in FIGS. 5 and 6 can be used in the manufacturing process of sputter targets. As mentioned above, the cleanliness of a sputter target is one factor that determines the quality of the layer or film produced by the target. Sputter target blanks are formed by molding only sputter target blanks having a cleanliness factor smaller than the standard cleanliness factor, in particular C _FN , or a histogram of selected columns or areas smaller than the standard value, and these conditions are not met Rejecting a blank can increase the likelihood that a sputter target so manufactured will produce a high quality layer or film.

Although the method described herein, and the form of apparatus for carrying out the method, constitutes a preferred embodiment of the invention, the invention is limited to the details of form of the method and apparatus. It is to be understood that neither can be made and both can be modified without departing from the scope of the invention defined in the claims.

[Brief description of drawings]

[Figure 1]   FIG. 3 is a schematic diagram showing a prior art method for ultrasonic tissue analysis.

FIG. 2 is a perspective view of a particularly preferred test sample before compression used for ultrasonic cleanliness characterization according to the present invention.

FIG. 3 is a schematic diagram showing a particularly preferable ultrasonic cleanliness characterization method using the compressed test sample shown in FIG. 2 according to the present invention.

[Figure 4]   FIG. 4 is a schematic diagram of a test apparatus for carrying out the method of FIG. 3.

FIG. 5 shows a proportion of “clean” Al-0.5 wt% according to a particularly preferred embodiment of the invention.
It is a histogram which shows the characteristic of Cu material.

FIG. 6 is a “cleaner” Al- than the above, according to a particularly preferred embodiment of the invention.
6 is a histogram showing the characteristics of a 0.5 wt% Cu material.

[Explanation of symbols]

  50 cylindrical sample   52 Disc-shaped test sample   54 Upper surface   56 Lower surface   60 transducer   62 ultrasonic pulse   64 echo   70 dimensions   72 defects   80 Immersion tank   82 XY scanner   84 controller   86 Band where ultrasonic pulse is focused   90 preamplifier   92 Linear amplifier   94 AD converter   95 Digital Oscilloscope   100 microprocessors   155 Horizontal axis   157 Vertical axis   161 Horizontal axis   163 vertical axis

Claims (33)

[Claims] 1. A non-destructive method for characterizing a sputter target material, comprising: a) using a reference sample to obtain a reference value for wavy phase inversion defects and a reference value for wavy phase non-inversion defects. B) irradiating the test sample with acoustic energy sequentially at a plurality of positions on the surface of the test sample of the sputter target material, c) detecting a high frequency echo waveform induced by the acoustic energy, and d) background texture. Distinguishing backscattering noise of interest from the echo waveform to obtain a high frequency echo waveform signal, e) monitoring the high frequency echo waveform signal for 180 ° waveform phase inversion, and f) high frequencies associated with each of the plurality of positions. The echo waveform signal is compared with the waveform phase inversion reference value and the waveform phase non-inversion reference value to obtain a void defect, a particle defect, and a defect-free defect. Give respective data points response, g) waveform phase inversion defects, counts the data point corresponding to the waveform phase inverting defects, and defect-free waveform phase number of defects counted corresponds to the inverted defective C _FI, waveform phase The defect count number C _FN corresponding to the non-inversion defect, the total defect count C _FT , and the total number of data points C _DP are determined. H) Total cleanliness factor F _CT = (C _FT / C _DP ) × 10 ⁶ , phase inversion Calculate the cleanliness factor F _CI = (C _FI / C _DP ) × 10 ⁶ corresponding to the defects and the cleanliness factor F _C = (C _FN / C _DP ) × 10 ⁶ corresponding to the non-phase inversion defects, each A method comprising steps. 2. The non-destructive method for characterization of sputter target materials according to claim 1, wherein the reference sample has flat bottom blind holes of constant depth and diameter. 3. A non-destructive method for characterizing a sputter target material according to claim 1, wherein the reference sample has intentionally embedded alumina particles of a specific size. 4. A non-destructive method for characterizing a sputter target material according to claim 1, wherein the test sample is a cylindrical body of the sputter target material. 5. A non-destructive method for characterizing a sputter target material according to claim 4, wherein the cylindrical body is molded into a disc-shaped test sample. 6. The non-destructive method for characterizing a sputter target material according to claim 5, wherein the disk-shaped test sample is formed by rolling the cylindrical object. 7. The non-destructive method for characterizing a sputter target material according to claim 5, wherein the disk-shaped test sample is formed by forging of the cylindrical object. 8. The non-destructive method for characterizing a sputter target material of claim 5, wherein the disk-shaped test sample has first and second flat surfaces. 9. The non-destructive method for characterization of sputter target materials according to claim 8, wherein the first and second flat surfaces are made by diamond cutter cutting. 10. The non-destructive method for characterizing a sputter target material according to claim 1, wherein the acoustic energy is generated by a transducer. 11. The test sample is further immersed in deionized water in a dip tank to cause the transducer to cause the acoustic energy to propagate through the deionized water such that the test sample is substantially on the upper surface of the test sample. The non-destructive method for characterizing a sputter target material according to claim 10, comprising the step of orienting so as to hit the test sample vertically. 12. The transducer is a piezoelectric transducer,
Has a fixed focal length in water of about 114 mm, a peak frequency of about 9.15 MHz with a bandwidth of about 8 MHz (-6 dB), the transducer having a focal area of about 21 mm in aluminum (-6
A non-destructive method for the characterization of sputter target material according to claim 10, characterized in that it produces pulses having a focal spot with a dB of 0.8-0.9 mm. 13. The test sample has an upper surface with a width of about 28 cm, and the entire upper surface has a thickness of about 0.
A non-destructive method for the characterization of sputter target materials according to claim 12, characterized in that the data points are acquired by a stepwise movement such as a raster scan in 9 mm steps. 14. A non-destructive method for characterizing a sputter target material, the method comprising: a) defining a plurality of amplitude signal bands for corrugated phase inversion defects; and b) a plurality of amplitude signal bands for corrugated phase non-inverting defects. And c) irradiating the test sample of the sputter target material with acoustic energy at a plurality of positions on the surface of the test sample sequentially to the test sample of the sputter target material, and the high frequency induced by the acoustic energy. Obtaining an echo waveform, d) conditioning the high frequency echo waveform signal to obtain a modified amplitude signal, e) measuring the modified amplitude signal to obtain a plurality of modified amplitude signal intensities, and f) the plurality of modifications. Comparing an amplitude signal strength to the plurality of amplitude signal bands to define a subset of the strengths of the modified amplitude signal; g) the portion of the modified amplitude signal strength. Counting the set to obtain a plurality of modified amplitude signal strength counts, and making each modified amplitude signal strength count of the plurality of amplitude signal strength counts correspond to one of the plurality of amplitude signal bands; h) waveform Creating a Pareto histogram for the phase inversion defect, i) creating a Pareto histogram for the waveform phase non-inverting defect, j) combining the histogram for the waveform phase inversion defect and the histogram for the waveform phase non-inverting defect, k) the modified amplitude signal A method comprising associating an intensity count number with the plurality of amplitude signal bands. 15. The conditioning of the high frequency echo waveform signal is performed by amplifying and filtering the high frequency echo waveform signal to produce the modified amplitude signal. A non-destructive method for the characterization of sputter target materials as described. 16. A non-destructive method for characterizing a sputter target material according to claim 14, wherein the modified amplitude signal is digitized. 17. A non-destructive method for characterizing a sputter target material according to claim 14, wherein the cylindrical body is molded into a disc-shaped test sample. 18. The non-destructive method for characterizing a sputter target material according to claim 17, wherein the disk-shaped test sample is formed by rolling the cylindrical object. 19. A non-destructive method for characterizing a sputter target material according to claim 17, wherein the disk-shaped test sample is formed by forging the cylindrical object. 20. The non-destructive method for characterizing a sputter target material according to claim 17, wherein the disk-shaped test sample has first and second flat surfaces. 21. The non-destructive method for characterizing a sputter target of claim 20, wherein the first and second flat surfaces are created by diamond cutter cutting. 22. The non-destructive method for characterizing a sputter target material of claim 14, wherein the acoustic energy is generated by a transducer. 23. Immersing the test sample in deionized water in a dip tank and causing the transducer to propagate the acoustic energy through the deionized water substantially perpendicular to the upper surface of the test sample. 23. A non-destructive method for characterizing a sputter target material according to claim 22, characterized in that it is oriented such that it hits the test sample. 24. The transducer is a piezoelectric transducer,
Has a fixed focal length in water of about 114 mm, a peak frequency of about 9.15 MHz with a bandwidth of about 8 MHz (-6 dB), the transducer having a focal area of about 21 mm in aluminum (-6
A non-destructive method for the characterization of sputter target material according to claim 22, characterized in that it produces pulses having a focal spot of dB) and a diameter of 0.8-0.9 mm. 25. The test sample has an upper surface with a width of about 28 cm, and the data points are moved stepwise such as a raster scan in steps of about 0.9 mm in both the x and y directions. 25. A non-destructive method for characterizing a sputter target material according to claim 24. 26. A transducer electrically connected to the echo acquisition system, the transducer being associated with both the waveform phase signal and the waveform amplitude signal.
An apparatus for nondestructively determining the characteristics of a sputter target material. 27. An apparatus for non-destructively determining the properties of sputter target materials according to claim 26, wherein said transducer is a piezoelectric transducer having a fixed focal length in water of about 114 mm. 28. The transducer is about 21 mm in aluminum.
28. Non-destructively characterizing the properties of a sputter target material according to claim 27, characterized in that it produces a pulse having a focal region (-6 dB) of 10 and a focal spot of 0.8-0.9 mm in diameter. apparatus. 29. The echo acquisition system comprises a low noise receiver connected to the transducer, the low noise receiver comprising a set of calibrated attenuators, 12 bits, 2.44 mV /
26. A non-destructive property of sputter target material according to claim 25, comprising a low noise linear amplifier having a bit AD converter and a digital oscilloscope connected to the analog output of the low noise receiver. Device to determine. 30. A mechanical XY scanner, further comprising a mechanical XY scanner electrically connected to the controller, wherein the transducer is attached to the scanner by specially configured attachment means. Item 27. An apparatus for nondestructively determining the characteristics of a sputter target material. 31. The controller is programmable, the controller having a program for controlling the mechanical XY scanner, data acquisition, and storage of acquired data points. Item 31. An apparatus for nondestructively determining the characteristics of a sputter target material. 32. The program of the controller is configured such that the defect count number C _FI corresponding to the waveform phase inversion defect, the defect count number C _FN corresponding to the waveform phase non-inversion defect, the total defect count number C _FT , and the data point Total number C _DP , total cleanliness factor F _CT = (C _FT /
C _DP ) × 10 ⁶ , cleanliness factor corresponding to phase inversion defect F _CI = (C _FI / C _DP
) × 10 ⁶ , cleanliness factor F _C = (C _FN / C _DP ) × corresponding to non-phase inversion defect
An apparatus for nondestructively determining the properties of a sputter target according to claim 31, characterized in that it calculates 10 ⁶ . 33. The program of the controller forms a Pareto histogram for a waveform phase inversion defect and a Pareto histogram for a waveform phase non-inversion defect, and combines the respective histograms for the waveform phase inversion defect and the waveform phase non-inversion defect. The apparatus for nondestructively determining the properties of a sputter target material according to claim 31, wherein the modified amplitude signal intensity counts are related to a plurality of amplitude signal bands.